Response to Reviewer# 1 2 (Q1) A scope of negative result is unclear

Thank you for helpful comments and suggestions. We will address the concerns raised by the reviewers. We discuss the difficulty to satisfy Corollary 5 when the problem becomes multiclass in Lines 144-158. This makes it easier to tune the hyperparameters to satisfy the necessary condition as illustrated in Eq. (9) for We checked Eq. (5) in "Learning Confidence for Out-of-Distribution Detection in Neural Networks" Regarding the problem addressed by "On calibration of modern neural